﻿Changes 
  of 
  Pressure 
  which 
  accompany 
  Point 
  Discharge, 
  25 
  

  

  minute 
  traces 
  of 
  oxygen 
  or 
  air 
  sending 
  N 
  up 
  to 
  over 
  100 
  for 
  

   negative 
  discharge. 
  But 
  there 
  seemed 
  to 
  be 
  a 
  lower 
  limit 
  to 
  

   the 
  contraction 
  in 
  the 
  purest 
  hydrogen 
  which 
  roughly 
  corre- 
  

   sponded 
  to 
  N 
  = 
  l. 
  This 
  was 
  for 
  a 
  copper 
  plate 
  *. 
  

  

  As 
  will 
  be 
  shown 
  further 
  on, 
  there 
  are 
  reasons 
  for 
  thinking 
  

   that 
  this 
  limiting 
  value 
  may 
  really 
  be 
  due 
  to 
  the 
  effect 
  we 
  

   were 
  looking 
  for. 
  To 
  be 
  sure 
  of 
  this 
  it 
  was 
  necessary 
  first 
  

   to 
  make 
  certain 
  that 
  impurity 
  was 
  not 
  its 
  cause 
  ; 
  and 
  we 
  

   therefore 
  studied 
  in 
  some 
  detail 
  the 
  effects 
  on 
  p 
  of 
  small 
  

   quantities 
  of 
  oxygen 
  and 
  nitrogen 
  in 
  the 
  hydrogen. 
  The 
  

   following 
  is 
  a 
  preliminary 
  account 
  of 
  this 
  work. 
  

  

  Method 
  of 
  the 
  Experiments. 
  

  

  In 
  fig. 
  1 
  is 
  shown 
  a 
  section 
  of 
  the 
  discharge 
  vessel. 
  PP' 
  are 
  

   discharge 
  points 
  of 
  the 
  finest 
  procurable 
  platinum 
  wire, 
  

  

  Fiff. 
  1. 
  

  

  sheathed 
  in 
  glass 
  tubes 
  to 
  within 
  a 
  few 
  millimetres 
  of 
  their 
  

   extremities 
  ; 
  QQ' 
  brass 
  blocks 
  to 
  the 
  flat 
  upper 
  surfaces 
  of 
  

   which 
  plates 
  of 
  various 
  metals 
  could 
  be 
  soldered. 
  P 
  and 
  P' 
  

   were 
  not 
  of 
  course 
  used 
  at 
  the 
  same 
  time. 
  

  

  To 
  reduce 
  the 
  warming 
  effects 
  of 
  the 
  current 
  as 
  much 
  as 
  

   possible, 
  the 
  discharge 
  vessel 
  was 
  formed 
  of 
  a 
  block 
  of 
  

   copper 
  AA, 
  10 
  x 
  12*5 
  x 
  15 
  cub. 
  cms., 
  in 
  which 
  were 
  two 
  

   platinum-lined 
  cavities 
  BB, 
  7*2 
  cm. 
  deep 
  and 
  5*0 
  cm. 
  in 
  

   diameter, 
  connected 
  respectively 
  to 
  opposite 
  ends 
  of 
  a 
  sensitive 
  

   tilting 
  pressure-gauge 
  capable 
  of 
  indicating 
  changes 
  of 
  pres- 
  

   sure 
  down 
  to 
  about 
  10 
  -8 
  atmosphere 
  f 
  . 
  

  

  Heat 
  produced 
  in 
  one 
  of 
  the 
  cavities 
  was 
  thus 
  after 
  a 
  short 
  

  

  * 
  With 
  platinum 
  and 
  palladium 
  plates 
  the 
  case 
  is 
  apparently 
  com- 
  

   plicated 
  "by 
  the 
  evolution 
  of 
  gas 
  which, 
  under 
  suitable 
  circumstances, 
  

   may 
  turn 
  the 
  contraction 
  into 
  an 
  expansion. 
  

  

  t 
  An 
  account 
  of 
  this 
  gauge 
  is 
  to 
  he 
  published 
  shortly 
  in 
  connexion 
  

   with 
  a 
  discussion 
  of 
  the 
  design 
  of 
  instruments 
  of 
  this 
  type, 
  

  

  